EMPLOYEE ORGANIZATION IN SILICON VALLEY: NETWORKS, ETI4C ORGANIZATION, AND NEW UNIONS

Alan Hydet

Was the software and marketing company that I shall call Individuate one of America's one hundred best places to work in early 2000?' Its human resources director, whom I shall call Farah, certainly thought it could be, and so did many of its employees. Its analytic software, which analyzed customer behavior during online and offline shopping, was very successful. Individuate had successfully merged two very different companies to develop the product. Originally a San Francisco marketing company catering to retailers, Farah told me later, it was full of "hip, stylish people from marketing and advertising, dressed in black." In 1999, they realized that they needed to merge with a database company and chose an East Bay firm named after a character from Egyptian mythology. The database company was "full of kids wearing t-shirts and jeans" and snobbish about their degrees from M.I.T., Cal Tech, and Stanford. Nevertheless, the merger had worked out great-only two of the East Bay kids had quit, because they did not want to commute into the city-and after the get-acquainted parties, Bay cruise, and community-building, everybody now "loved the new company." Business was great, salaries were high, and benefits were generous-including stock options, full medical and dental coverage, a 401(k) plan, free snacks and drinks, a recreation room, and reduced rates on gym membership. Like every other information technology or software company in the Bay Area that year, the merged company was hiring all the time. Executive management thought that it might help hiring and retention to have the

t Professor and Sidney Reitman Scholar, Rutgers University School of Law, Newark. This article is adapted from Chapter Five of the book, WORKING IN SILICON VALLEY: ECONOMIC AND LEGAL ANALYSIS OF A HIGH-VELOCITY LABOR MARKET, to be published in 2002 by M.E. Sharpe, Inc. A. Hyun Rich provided research and editorial assistance in the preparation of this Article. 1. The Individuate story comes from an interview with its former personnel director on August 1, 2001. The company's name, the personnel director's name, and many details about the company have been changed. 494 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3 company recognized widely as a great employer, perhaps by being named a "top company to work for" by business media. Farah looked into it and learned that to make these lists, Individuate would have to show that it was genuinely meeting employee demands. Therefore, Farah called open meetings to find out what the staff wanted. The chief demand was for further professional development and training. "The engineers wanted classes in software skills. The sales people and customer support staff wanted to know more about engineering." A small volunteer group formed to work up a plan, and Farah met with them weekly. Consultants had warned that it would be dangerous to raise expectations, but Farah was "green," having only two years of experience as the human resources director of the East Bay database company. The volunteer group also realized that they wanted more information about the company's finances and less hype. This became known around Individuate as "respect." After all, the employees, at least at the time of the merger, had just worked for a company in which they saw the CEO daily and asked him or her whatever they wanted. A new Individuate CEO in early 2000 was exploring expanding and taking the company public, and the employee groups wanted more information about what this meant. Although management had thought it would be "cool" to be one of the hundred best places to work in America, it never met these employee concerns. The company's stock fell, and management time was devoted to the basic objective of steering the company through a rough market. Employee initiatives, like training or sharing financial information, seemed like a diversion. Management even opposed a suggestion box for anonymous suggestions. More importantly, people felt demoralized and began to take jobs elsewhere, easily done in mid-2000. By then, demand had peaked for Individuate's software. The company stopped hiring in late 2000 and downsized in early 2001. Its plans to go public have been shelved. The lesson of the Individuate story seems ambiguous to me. The company seems typical. Its only unusual feature was its goal to be listed as one of the "hundred best companies to work for." Not many companies engaged in such projects. Nevertheless, plenty of Silicon Valley companies in 2000 probably believed that they belonged on that list. I think that most of them, had they pursued that listing, would have discovered, as the consultants warned Farah, that introducing employee consultation into an organization that lacks responsiveness to employee suggestions can create false expectations. What is the solution to this problem? Does the Individuate story show the need for new forms of employee organization in Silicon Valley? Or does it demonstrate the need for old-fashioned employee organization like labor unions? Does it indicate the inevitability of informal modes of representation such as 2002] EMPLOYEE ORGANIZATION IN SILICON VALLEY meetings, volunteer committees, and intranet communication? Or does it merely demonstrate the harm that comes from consulting employees at all? Ambiguity about formal employee organization pervades much of the literature on high-tech employment. The Silicon Valley engineer or manager has symbolized both social isolation and a high level of networking. Scenes of Bill Hewlett, David Packard, Steve Jobs, and Steve Wozniak tinkering in garages update the image of Charles Lindbergh individualism. Similarly, the first character introduced in Robert Bellah's influential book is a Silicon Valley executive named Brian Palmer.2 He stands for devotion to work and family, commitment to individualism, and isolation from any kind of political or social organization.3 By contrast, the managers and engineers in AnnaLee Saxenian's book pushed Silicon Valley ahead of Boston's Route 128 due to their dense network of personal ties with their counterparts at rivals, subcontractors, suppliers, and former employers.4 Nevertheless, neither the isolate nor the networked seem to have much time for formal group membership. Information-labor markets create roles for new labor market intermediaries, like Internet job boards, which match employees to jobs and provide information.5 Do these intermediaries create opportunities for employee organizations, such as old-fashioned unions or some new kind of employee organization? Many people have speculated along these lines or worked to create such new organizations, but so far results are thin on the ground. This article will discuss four kinds of employee organizations and how they relate to information-labor markets like Silicon Valley's. They include network-based employee groups that communicate on computer networks; ethnic and gender-based groups that transcend firm boundaries; other groups organized as caucuses by firms; and new employee groups designed for mobile workers, such as Working Partnerships in San Jose, and Working Today, a New York-based group not yet active in Silicon Valley. These new forms of employee organizations are full of vitality, play

2. See ROBERT N. BELLAH ET AL., HABITS OF THE HEART: INDIVIDUALISM AND COMMITMENT IN AMERICAN LIFE 3-8 (1985) (introducing the Brian Palmer character). 3. Id. 4. See generally ANNALEE SAXENIAN, REGIONAL ADVANTAGE: CULTURE AND COMPETITION IN SILICON VALLEY AND ROUTE 128 (1994) (contrasting the industrial systems between Silicon Valley and Route 128). 5. See David H. Autor, Wiring the LaborMarket, 15 J. EcON. PERSP. 25, 26-30 (2001) (discussing the consequences of the Internet for labor markets). HYDE, supra note *, at ch. 4, discusses the implications of such new labor market intermediaries. The premise of an "information-labor market" is that labor markets with short tenures and weak internal labor markets are markets for information. Individuals may be hired for the information they contribute. One sees network effects, endogenous growth through information spillover, multiple equilibria, increasing returns, as well as start-up firms, job turnover, information intermediaries, flexible compensation. Id. 496 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3 interesting roles, and raise policy dilemmas that are best resolved by examining the full picture of this high-velocity labor market. Like other Silicon Valley work practices, they also raise questions at the cutting edge of economic analysis.6 Specifically, they all are institutions of information transmission, also known as "employee voice." They have little bargaining or other labor market power. The economics of institutions of employee voice are not well understood. Yet whether any of them will endure in the new economy is somewhat doubtful, particularly as compared with Internet job boards and websites, and temporary help agencies.7 The question that pervades this entire article is whether an employee-directed organization, which seeks to provide assistance, support, and information to employees, offers any advantage over proprietary organizations that aspire to provide the same services. For lower-compensated employees in the labor market, such as office temps, and perhaps even programmers, none of these new groups is a good substitute for a traditional union. If unionism ever comes to Silicon Valley's high-velocity labor market, it will likely more closely resemble the old-fashioned American construction union. The construction union is the best organization ever devised for a mobile or contingent workforce- so good that one rarely thinks of unionized construction workers as "contingent." Today's temps, product-testers, and low-level programmers, would be much better off with an organization that administered (jointly with employer groups) funds, that paid retirement, health, and vacation benefits, into which employers paid per hour worked, organizations that also trained and certified employees. Since, as we shall see, current labor law makes it unnecessarily difficult for temporary and mobile workers to form unions, that law should be changed. If unions were successful in such labor markets, they could in turn be the model for higher-salaried technical labor like customer service representatives or even programmers. This article will describe such an organization after surveying the forms of employee organization that actually exist in the Valley. First, this article will examine existing unions.

I. UNIONS

Rates of union membership are not available by county, but most people think that Santa Clara County is very lightly unionized. In high- technology industries, unions are hardly present at all. At a high- technology company, the only employee who is typically represented by a union is the janitor, most likely represented by Local 1877 of the Service

6. See HYDE, supra note *, at ch. 4 (analyzing a high-technology labor market as a market for information). 7. See id. 2002] EMPLOYEE ORGANIZATION IN SILICON VALLEY

Employees International Union.8 Recruiting such a janitor is a remarkable organizing feat. Almost all the usual excuses for failed union organizing might have applied to that janitor. He or she is probably an immigrant from Mexico with little English language proficiency. The janitor has few job skills and can easily be replaced. Finally, the janitor is not an employee of a high-technology company like Sun or Apple, but typically works for a cleaning contractor. The Service Employees became the janitors' union after a brilliant organizing campaign built around appeals to Mexican cultural pride. The campaign involved religious and political leaders, demonstrations, and fiestas. The coup de grace was a consumer boycott of Apple computers and other products sold by high-technology companies that hired the cleaning contractors.9 Local 1877 now represents a majority of the Valley's janitors.0 Besides Local 1877, membership in labor unions in high-technology companies in Silicon Valley is minimal. The Communication Workers made news headlines in late 2000 with organizational efforts to form a union among customer service representatives at the San Francisco on-line retailers etown and Amazon.com in Seattle." This union organizing effort ended when etown went out of business and Amazon.com eliminated the corporate division in which the union had the greatest strength.1 2 In the 1970's, when there were more manufacturing companies in the Valley, there were unsuccessful union attempts to organize production workers. 3 The only known unionized Web business is a special case. Online Television Network Services, Inc. (OTVnet.com) is a San Diego firm that develops on-line communication tools for labor union clients administering

8. Gary Rivlin, Cleaning Up in Silicon Valley, INDUSTRY STANDARD, Apr. 24, 2000, availableat http://www.thestandard.comlarticle/0,1902,14319,00.html. 9. Christian Zlolniski, The Informal Economy in an Advanced IndustrializedSociety: Mexican Immigrant Labor in Silicon Valley, 103 YALE L.J. 2305, 2313-16 (1994); Jesis Martinez Saldaiia, At the Periphery of Democracy: The Binational Politics of Mexican Immigrants in Silicon Valley 115-25 (1993) (unpublished Ph.D. dissertation, University of California, Berkeley). See generally David Bacon, Organizing Silicon Valley's High Tech Workers (1997), at http:lwww.igc.orgldbaconlUnions/O4hitecO.htm (last visited Nov. 26, 2001) (discussing union organizing across several industries). The consumer boycotts did not violate federal prohibitions on secondary boycotts, which do not reach consumer boycotts carried out by means other than picketing. See Edward J. DeBartolo Corp. v. Fla. Gulf Coast Bldg. & Constr. Trades Council, 485 U.S. 568, 569 (1988) (holding that even if handbilling not involving a distributor is outside the publicity proviso's protection, that does not require the conclusion that distributing handbills is coercive under § 8(b)(4)(ii)). 10. Rivlin, supra note 8. 11. David M. Ewalt, IT Workers Look at the Union Label - Dissatisfied Dot-corn Workers Look to Labor Organizers to Blunt Effects of Economic Slowdown, INFO. WK., Mar. 19, 2001, at93. 12. Id. 13. Bacon, supra note 9, at 10. 498 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3 multi-employer benefits plans. Its management was happy to recognize Local 569 of the International Brotherhood of Electrical Workers as the representative of its five production and development technicians. 4 Their collective bargaining agreement prohibits discharge without cause and offers unusual benefits. 15 Almost all the work practices identified with Silicon Valley, which are discussed more fully in Working in Silicon Valley, are said to be obstacles to union organizing, particularly short job tenures, heavy use of temporary labor, and heavy use of immigrant labor. 6 But, as the example of the janitors' union suggests, these characteristics may also present opportunities for union organizing. Union image is often the biggest obstacle to union organizing. When unions come up in my interviews, people in the Valley, who rarely have much direct experience with unions, associate unions with traditional union roles in traditional labor markets. Many people who respond this way are thinking of real union practices that, like the economics and law of employment relationships, have developed against an assumption of internal labor markets, typically in manufacturing industries. Such unions negotiate detailed multi-year collective agreements that link compensation and benefits to detailed job descriptions, with each job holding a place in a defined hierarchy. The benefits included in such agreements are back- loaded, creating returns for seniority. Management agreed to these terms to discourage turnover and realize other gains that come from a stable, senior workforce. 17 Companies that do not gain the benefit of a senior workforce may resist unions to such a degree that they discourage employee interest.18 In such a case, the union's only real weapon is to strike. Strikes expose employees to the risks of job loss and to the stress of conflict, and may only succeed against some employers. This image of unions is at best a partial picture of how unions work. Such an image is what cognitive scientists call a prototype.' 9 For example, if asked to picture a bird, most people in our culture will picture a robin or bluebird. Few will imagine an ostrich or flamingo. Similarly, the prototype for unions for years has been the United Automobile Workers. When asked to picture a union, people automatically think of automobile workers. People also picture union work, detailed contracts, job

14. John Roemer, Should Tech Workers Unite? INDUSTRY STANDARD, Jan. 23, 2001, at http://www.thestandard.com/search?keyword=roemer &stype=d&tf= E& month=01 &day= 23&year=-2001 #newsletter. 15. Id. 16. See HYDE supra note * 17. See generally RICHARD B. FREEMAN & JAMES L. MEDOFF, WHAT Do UNIONS Do? ch. 8 (1984) (asserting that unions create returns for seniority). 18. RICHARD B. FREEMAN& JOEL ROGERS, WHAT WORKERS WANT ch. 4 (1999). 19. STEVEN L. WINTER, A CLEARING IN THE FOREST: LAW, LIFE AND MIND (2001). 2002] EMPLOYEE ORGANIZATION IN SILICON VALLEY

descriptions, formal grievance systems, and fixed compensation. This image of unions offers little to most employees in high-velocity labor markets and even less to their employers. There is currently very little manufacturing of any kind in Silicon Valley.20 Before the 1990's, large firms like Hewlett-Packard and Sun did manufacturing work in Silicon Valley. During this period, no new unions emerged. Manufacturing work was largely conducted by non-unionized immigrant women. By now, most manufacturing has moved overseas and is no longer carried out by the "manufacturers" of computers or equipment, who outsource over eighty percent of production work.22 Industrial unionism is not the only kind of unionism. Employees in Silicon Valley have developed some of their own forms of organization. This article discusses employee groups on electronic networks, ethnic and gender based networks and formal caucuses, and some new associations designed specifically to appeal to mobile, educated workers. On the one hand, unions may adopt these organizational forms, possibly in combination with some traditional union forms. On the other hand, unions may never become established in Silicon Valley, but the new organizations may flourish. Neither of these outcomes is likely. Neither the Valley, nor any other portion of the U.S. economy, will likely ever become heavily unionized, and the organizations in this article-networks, caucuses, and new employee groups-are unlikely to involve more than small pockets of the labor force. However, the unconventional nature of these groups has already challenged traditional labor and employment law. It is therefore important to examine how the legal system responds to these challenges brought by new forms of employee organization. Employees and employers in Silicon Valley often associate labor unions with traditional labor markets, and therefore view them as antithetical to their high-velocity labor market. One of the most robust generalizations in industrial relations, however, is that unions take their structure from the employers they face. In truth, they extrapolate the

20. Dieter Ernst, FROM PARTIAL TO SYSTEMIC GLOBALIZATION: INTERNATIONAL PRODUCTION NETwORKs IN THE ELECTRONICS INDUSTRY (Berkeley Roundtable on the Int'l Econ., Univ. of California, Working Paper No. 98, 1997), available at http:l/vww.ciaonet.orglwps/erdol (last visited Nov. 29, 2001). 21. See generally GUADALUPE M. FRIAZ, LABOR STRATIFICATION AND DOWNSIZING IN COMPUTER MANUFACTURING: EFFEcTS ON WHITE WOMEN AND WOMEN OF COLOR (Julian Samora Research Inst., Working Paper No. 23, 1996) (demonstrating that women held a large percent of the computer manufacturing jobs); Karen J. Hossfeld, 'Their Logic Against Them': Contradictionsin Sex, Race, and Class in Silicon Valley, in WOMEN WORKERS AND GLOBAL RESTRUCTURING (Kathryn Ward, ed., 1990); Bacon, supra note 9 (analyzing the difficulties of unionizing in Silicon Valley and concluding that labor laws are frequently not helpful in protecting workers in this high-tech region). 22. Ernst, supra note 20. 500 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3 problems they address from specific employers. 23 For instance, unions that face small craft-oriented firms will organize on craft lines. Some European or Latin American unions that face large corporate national enterprise, largely funded by the national capital under government guidance, focus their attentions on political parties and centralized bargaining. The economic story of the twentieth century was, until recently, the rise of the large corporation that realized economies of scope and scale.14 Unions that deal with such corporations are associated with detailed contracts, job descriptions, and seniority rights. None of these features, however, is inherent to labor unions, or would have much relevance in a high-velocity labor market. Therefore, there is a need to examine new high-velocity labor markets against a scholarly tradition in economics, law, and industrial relations that so often assumes the stable career ladders in large corporations.

II. EMPLOYEE ACTION ON COMPUTER NETWORKS

Employees with access to in-house computer networks can take collective action with a speed and efficacy that many traditional unions might envy. Consider the action by employees at Apple Computer in 1990 described by Bishop and Levine.25 Apple had encouraged employee use of an internal computer bulletin board. Its training materials noted that its recycling program "got started when someone suggested it in a [network] discussion. 2 6 Other discussions had concerned policies on hiring minorities, smoking in buildings, and naming conventions for shared computers. 27 Fifty top executives received periodic two-page summaries of the main issues discussed on the network, under the title "What Employees ,2 Want. 1 In January 1990, Apple management announced revisions to the profit-sharing plan that would have eliminated any payments to employees in quarters with slow sales growth.29 "Literally hundreds of postings were entered on [the electronic bulletin board] in the busiest days. All told, on

23. See generally CHARLES C. HECKSCHER, THE NEW UNIONISM: EMPLOYEE INVOLVEMENT IN THE CHANGING CORPORATION (1988) (arguing that unions borrow their organizational structure from employers). 24. See generally ALFRED D. CHANDLER, JR., SCALE AND SCOPE: THE DYNAMICS OF INDUSTRIAL CAPITALISM (1990) (describing the development of the managerial business enterprise that provides the central dynamic to global capitalism). 25. See Libby Bishop & David I. Levine, Computer-Mediated Communication as Employee Voice: A Case Study, 52 INDUS. & LAB. REL. REV. 213, 218-24 (1998). 26. Id. 27. Id. 28. Id. 29. Id. 20021 EMPLOYEE ORGANIZATION IN SILICON VALLEY the order of one thousand messages were received. This issue elicited a greater volume of response than had any event in the history of the company. The system was literally swamped. 30 Most responses attacked the reductions in the profit-sharing payments, often bitterly, while a 3 minority defended them. ' A common early theme was that "employees wanted a complete explanation for why the change in profit-sharing was32 necessary, and how it would encourage growth as management claimed., CEO John Sculley posted such explanations on the bulletin board, but many employees remained unconvinced. Ultimately, management caved in. Everyone interviewed for a Harvard Business School case attributed this to the effect of the electronic bulletin board.33 Meanwhile, one employee looked through bulletin board transcripts and recorded all the contributors to the profit-sharing debate who had used their actual names. He wrote to each identifiable contributor, suggesting that they should all meet. A small group formed and met quietly several times that winter. Independently, a second employee posted a bulletin board message calling for the formation of a concerned employees' league. The two groups later merged under the name "Employees for One Apple." In May 1990, fifty employees attended the first meeting of the merged group, which was announced on the bulletin board. The group sought the restoration of the "corporate culture they recalled (perhaps in a somewhat idealized form) from a few years previous." They sought increased communication with top management, fewer management "perks" and other distinctions among employees, and an institutionalized voice for employees. "Members of One Apple" repeatedly pointed out that they loved Apple. Conversely, most members and leaders we interviewed went out of their way to note that they were opposed to a union at Apple... they did not want outsiders involved and they did not want a union bureaucracy to intervene between employees and managers." 34 Management, however, resisted formal employee representation. As of the date of Bishop and Levine's study, the only effect of the agitation over organization was the formation of an Employee-Executive Forum in which fifteen employees, randomly selected from among volunteers, discussed their concerns with top managers.35 The Forum had no rights to information or consultation and no decision-making authority. It was supposed to meet quarterly but, in fact, met less frequently.

30. Id. 31. Id. 32. Id. 33. Michael J.Gibbs, Apple Computer (B): Managing Morale and CorporateCulture, 9 HARV. Bus. SCH. CASE491 (1991). 34. Bishop & Levine, supra note 25, at 221. 35. Bishop & Levine, supra note 25, at 223. 502 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3

Other companies that have announced changes in compensation, both inside and outside of Silicon Valley, have similarly faced what the New York Times calls "The Electronic Rank and File".36 In 1999, IBM reversed its previous position and more than doubled the number of employees permitted to keep a traditional, defined-benefit pension plan. The following year, Bell Atlantic similarly announced, then retracted, plans to switch employees to a "cash balance" plan with variable benefits. In each case, employees set up websites that compared the two plans and linked to the websites regarding pension rights published by government agencies, Congressional offices, and other organizations.37 The IBM site was visited more than 97,000 times that year and became a model for employees at other firms.38 "I just don't see how we could have done it without the Web," said Lynda French, creator of the cashpensions website.39 Although they offered no alternative explanation for their changes of mind, both IBM and AT&T denied that they had been influenced by the on-line pressure from employees.

A. Economic Analysis of Employee Voice Groups

Economic analysis of unions often begins, as a result of Freeman and Medoff's work,40 by distinguishing the effects of a cartel of labor from the effects of a "voice" for labor. The economic analysis of unions as labor cartels is well developed, while few generalizations can be made about the effects of union "voice." In the case of a standard, old-economy union, it may be very difficult to untangle the economic effects of its different functions. A unionized company, as compared with a nonunion counterpart, will pay higher wages and benefits, has a lower turnover rate, and will be more likely to terminate employees than adjust compensation. Do these effects, commonly associated with unions, reflect "cartels" or "voice?" The cartel effect is evident due to the union's ability to threaten or organize a strike. The voice effect is evident when the employer in a unionized workplace focuses on the average, as opposed to the marginal, employee. It might seem that this

36. Virginia Munger Kahn, The Electronic Rank and File, N.Y. TIMES, Mar. 8, 2000, at G1. 37. Id. 38. Id. 39. Kahn, supra note 36, at G1. See also Cash Balance Pensions and Retirement Plan Conversions Website!, at http://www.cashpensions.com (last visited Nov. 12, 2001) (comparing plans and providing information on pension benefits). 40. See generally FREEMAN & MEDOFF, supra note 17 (analyzing the role of trade unions in the United States by comparing outcomes in union settings to similar non-union settings and concluding that unions have both a monopoly and a voice "face" and that unions can improve the free enterprise system in the United States). 20021 EMPLOYEE ORGANIZATION IN SILICON VALLEY

voice effect would never be observable unless the union had its cartel power. Perhaps few employers would learn much about employee preferences from a union, given the more sophisticated tools available to learn about employee sentiments, including quality of work life groups and teams. Based on the assumption that employers are so well-informed, one would think that only the threat of economic harm through collective bargaining could change employers' behavior. This view is mistaken, for reasons explained below. Groups such as the networked groups at Apple or IBM present a novel problem for the economic analysis of institutions of employee "voice." They have no cartel effects, nor are cartel effects feasible in a high-velocity labor market. These groups do not strike. Their effect on turnover and "quits" is unknown. It may be significant that neither Apple nor IBM is a stereotypical startup with rapid turnover (such as Individuate). Neither research nor interviews turned up accounts of similar employee-networked action at a high-turnover startup. Surely there must have been such actions, at least over unhappiness about relatively minor job dissatisfactions. In a startup, though, such actions could not be based on the rhetoric of restoration of the former regime. Perhaps such rhetoric is more important to electronic action than it seems. Electronic action may be hard to organize except in defense of existing benefits. Most likely, employee action on electronic networks has a very bright future, far beyond the reported instances in which it has been mobilized in defense of existing benefits. Network mobilization is a spectacular vehicle for employee voice, which can be organized cheaply, quickly, and effectively. Management has conceded under pressure from these organizations, even when it did not face a threatened strike. The Apple employees who participated in network mobilization did not understand the extent of their own power. It is ironic that so much employee follow-up went into creating silly and ineffective new institutions of employee voice. These employees had already demonstrated their command over an effective institution of employee voice-one that mobilizes not only employee complaints, but managerial responses. A deeper irony is that employees were effective without formal organization. When they were simply posting complaints on the network, management responded to the complaints and changed its course. After forming an organization and seeking permanent channels of communication, they accomplished little. While these Apple employees were not poor people, they would seem to represent another illustration of Piven and Cloward's thesis that popular protest often becomes ineffectual when organizers transform it into a mass permanent organization.41

41. See generally FRANCES F. PIVEN & RICHARD A. CLOWARD, POOR PEOPLE'S 504 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3

Of course, employee action on computer networks faces some severe limitations. So far, it has been exclusively reactive. It is hard to imagine how network groups could ever plan, link with other groups, or employ professional assistance. Nevertheless, as means of mobilizing employee voice for management response, it has a bright future, and will generate legal cases in addition to those discussed below.

B. Employee Voice and "Morale"

It is not surprising that there is no developed economic analysis of (pure) employee voice groups, since the concept appears for the first time here. Such economic analysis might begin, however, by focusing on employee morale, the striking and unexpected finding of Bewley.42 Bewley conducted over three hundred interviews during the Bush recession of the early 1990's in order to learn why employers are so much more likely to lay off employees than they are to adjust wages downward. Most subjects named employee morale, a factor difficult to model economically and often dismissed by economists. 43 Bewley's informants use "morale" to cover all the factors that motivate high employee performance without immediate reward and often without direct monitoring by managers. 44 "Employers want workers to operate autonomously, show initiative, use their imagination, and take on extra tasks not required by management; workers who are scared or dejected do not do these things. '45 Good morale includes employee satisfaction, a sense that effort will be rewarded, and the expectation that employees will cooperate spontaneously and share information among themselves and with supervisors. 46 The last two factors are particularly difficult to motivate either with threats or financial incentives.47 Pay cuts cause bad morale in several different ways. The best employees will leave. Others will shirk or retaliate. Reducing the compensation of new hires is not a good idea either. "Managers regarded anything that upsets internal equity as potentially disruptive. Lack of equity spawns jealousies, resentments, and perceptions of unjust treatment.,as

MOVEMENTS: WHY THEY SUCCEED, How THEY FAIL (Vintage Books ed., 1979). 42. See generally TRUMAN F. BEwLEY, WHY WAGES DON'T FALL DURING A RECESSION (1999) (analyzing wage rigidity during economic downturns by surveying business consultants in the Northeast during the recession of the early 1990's and concluding that complicated employee behavior leads to manager reluctance to cut pay). 43. Id. 44. Id. 45. Id. at 431. 46. Id. 47. Id. 48. Id. at 433. 2002] EMPLOYEE ORGANIZATION IN SILICON VALLEY

Bewley does not specifically discuss collective bargaining. There was a union presence in thirty-five of the 235 companies in his sample, but those companies rarely appeared to behave in any special way.49 Nevertheless, Bewley's book represents an interesting vantage point from which to rethink collective bargaining. It certainly captures the dynamic of the Apple, IBM, and AT&T stories better than conventional models of collective bargaining. Why did these employers give in to the disgruntled employees on the internal network, when those employees did not threaten a strike or boycott? When one understands the importance of what management calls "morale," one can understand how "voice" institutions might be effective without a strike threat. Management will go to great lengths to avoid low employee morale. Perceived injustice or inequity leads to bad morale. Institutions of employee voice that publicize potential injustice or inequity transmit information regarding these perceptions to employees and managers. They effect swift management response, lest the best employees leave or all employees slacken their voluntary cooperation. Once again, a problem in the economics of labor and employment turns out to be, to a significant degree, a problem in the economics of information. Employee organizations that have only information effects may, under some circumstances, be as effective or more effective than organizations that can threaten or inflict economic harm through a strike. Future models of even the prototypical version of collective bargaining will have to account more effectively for this potent informational effect. Networks would be appealing if they involved less than a majority of employees, or if the employer opposed union organization. Either or both of these situations exists in most U.S. workplaces, certainly including Silicon Valley. By contrast, if the employer does not oppose employee unionization, and a union represents a majority, it is hard to think of any advantage to the employees in maintaining a network as opposed to a union. If a union represented a majority of the employees in an appropriate bargaining unit, it would have legal privileges that a network would lack. The majority union could use the law to require employer bargaining, have dues withheld from employee paychecks, or impose certain limited discipline on its members. 50 A network with these privileges in addition to its rapid mobilization of employee voice and employer response, would, in theory, be more effective than a network without them. As Al Capone allegedly said: you can go further in life with a kind word and a gun than with a kind word alone. However, in practice, these legal privileges of a

49. See id. at 32-33 (discussing the wide variety of businesses studied, none of which stood out in a particular way). 50. National Labor Relations Act, §§ 8(a)(3), 8(a)(5), 8(b)(1)(A), 29 U.S.C. §§ 158(a)(3), 158(a)(5), and 158(b)(1)(A). 506 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3

union might cut against the power of employee networks by triggering the kind of employer opposition that networks have not so far faced. Employers do have legal ways of not cooperating with unions, such as refusing to recognize them until the government conducts an election, then campaigning against them. Many employees, like those surveyed in Freeman and Rogers, fear such employer opposition." Employees might be drawn to network organization, particularly if employees were more interested in employer responsiveness than in formal bargaining, since employers do not seem to fight network organization.

C. Legal Issues Raised by Employee Voice Groups

No new legal framework is necessary to accommodate the new network organizations of employee voice. These organizations are already protected by existing labor law, which gives them two major privileges: protection against employer retaliation, and some rights to equal treatment by the employer if the employer has a preferred employee organization. 52 These legal rights of such groups, and employees who participate in them, come from the federal National Labor Relations (Wagner) Act.53 The source of these rights is ironic because the employees involved in network organizations are always careful to point out that they do not like unions and are not trying to organize one. Nevertheless, the Wagner Act protects the right of employees to engage in "concerted activities for the purpose, ' 54 not merely of collective bargaining, but of "other mutual aid or protection. 5 If an employee were fired or otherwise punished for complaining about bonus changes on internal networks, or for trying to get other employees to do something about it, such an action would violate the Wagner Act and would be corrected by the National Labor Relations Board on complaint by the employee.5 The latter case specifically protects employee internal e-mails, cast in the sarcastic tone so often found in e-mail.57 If the employer has its own preferred forms of employee organization, it may have to extend the

51. See generally FREEMAN & ROGERS, supra note 18. 52. See PATRICIA WALLACE, THE PSYCHOLOGY OF THE INTERNET 110-32 (1999) (discussing aggression found on-line); Sara Kiesler & Lee Sproull, Group Decision Making and Communication Technology, 52 ORGANIZATIONAL BEHAv. & HUM. DECISION PROCESSES 96, 110 (1992) (showing that, unlike face-to-face communications, computer discussion is more rude and impulsive). 53. National Labor Relations (Wagner) Act, 29 U.S.C. §§ 151-169 (1994). 54. 29 U.S.C § 157. 55. Id. 56. See Timekeeping Systems, Inc., 323 N.L.R.B. 244, 250 (1997) (holding that employee remarks on internal e-mail systems are protected concerted activity under N.L.R.A. § 7). 57. See sources cited supra note 52. 2002] EMPLOYEE ORGANIZATION IN SILICON VALLEY same privileges to the employee-organized group.58 For example, at the time of the employee action on Apple's internal network, Apple promoted employee caucuses including Black and Asian caucuses. Apple let those groups have e-mail boxes, meeting spaces, and bulletin boards, and therefore was likely obligated to extend the same privileges to an employee network group,59 even if neither group is a statutory "labor organization. 60 There is no dispute over the proposition that employee action on computer networks, or other informal group action without a union, is nevertheless protected by federal labor law. There are, however, two significant areas of legal uncertainty surrounding this protection. First, the law on protected activity under the Wagner Act has often been applied to deny protection to common action by unorganized employees.6 This is particularly true when orders of the National Labor Relations Board come before federal appeals courts. Judges insist, incorrectly, that the existence of a union is necessary to obtain the Wagner Act's protection, or deny protection because of the Rabelaisian language often favored by non- unionized employees, particularly on e-mail. 62 Judges also invoke vague rules about protest going "too far."63 These holdings are likely a result of judicial hostility to the Wagner Act itself. However, the holdings also reflect, in part, the power of that prototype of union organizing(e.g. the robin, not the flamingo), and a refusal to understand that protecting the actual activity of unrepresented employees is an equal priority of the Wagner Act. This priority requires protecting the ways that non-unionized employees present their grievances, not merely those channels that seem appropriate to federal judges. The more difficult legal question concerns access to company e-mail systems. For example, Ken Hamidi is a former Intel engineer, bitter about his termination, who is the most prominent figure in Former And Current Intel Employees (FACE). FACE maintains a website posting information

58. See Alan Hyde et al., After Smyrna: Rights and Powers of Unions That Represent Less Than a Majority, 45 RuTGERS L. REv. 637, 659-63 (1993) (discussing how an employer must extend the same privileges to an insurgent group as those enjoyed by the employer's representation plan). 59. See Black Grievance Comm. v. NLRB, 749 F.2d 1072, 1072 (3d Cir. 1984) (holding that preferential treatment to one non-majority group over another is unlawful interference). 60. NLRB v. Northeastern Univ., 601 F.2d 1208, 1217 (1st Cir. 1979). 61. See Alan Hyde, Employee Caucus: A Key Institution in the Emerging System of Employment Law, 69 CHI.-KENT L. REv. 149, 171 (1993), reprinted in THE LEGAL FUTURE- OF EmpLOYEE REPRESENTATION (Matthew Finkin ed., 1994) (arguing that the Wagner Act needs to be interpreted so it can protect unorganized employees). 62. New River Indus. v. NLRB, 945 F.2d 1290 (4th Cir. 1991) (denying protection to Rabelaision language by unrepresented employees); Reef Indus. Inc. v. NLRB, 952 F.2d 830 (5th Cir. 1991) (protecting such language because of union support). 63. NLRB v. Motorola, Inc., 991 F.2d 278,283-85 (5th Cir. 1993). 508 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3 on lawsuits against Intel by former employees alleging wrongful termination, and other information critical of Intel.64 FACE is a kind of network group, as the term is defined in this article. An Intel employee could not be fired or disciplined for joining or participating in FACE. Hamidi raised the legal stakes of employee network use by sending e- mail messages to current Intel employees over Intel's internal network, an activity that a California court has enjoined as trespass. 65 "Trespass" may seem like a ridiculously old-fashioned way of thinking about an e-mail network or any system of communications. After all, Intel probably could not enjoin Hamidi from mailing letters to each of its employees. Nonetheless, "trespass" is a surprisingly important concept in labor law, and is turning out to be equally important in the law of electronic data. In labor law, the concept of trespass permits an employer to bar anyone who is not an employee from its physical property including union organizers who are not employees. While employees have a federally protected right to talk about organization among themselves and to hand out literature, the employer does not have to let union organizers or any other nonemployees onto its property. This right is framed as a property right. The employer does not need to have any reason, and it does not matter if its property is normally open to the public, such as a shopping center.66 There is, however, a dispute between the NLRB and some lower courts that the Supreme Court may eventually resolve, as to whether an employer that permits charitable solicitation of employees using company property must extend similar privileges to union organizers. 67 Finally, state courts may enjoin trespassers even if the NLRB might have held that they are protected in cases where the alleged trespasser did not make any effort to get the case before the NLRB.68 So, if the internal network is Intel

64. Former and Current Intel Employees, at http://www.faceintel.com. (last visited Mar. 19, 2002); see also BILL LESSARD & STEVE BALDWIN, NETSLAVES: TRUE TALES OF WORKING THE WEB 182-97 (2000) (fictionalizing an account of Hamidi's activities). 65. Intel Corp. v. Hamidi, 114 Cal. Rptr. 2d 244 (Cal. Ct. App. 2001), review granted and depublished, 2002 Cal. LEXIS 1883 (Cal. Mar. 27, 2002). 66. See Lechmere, Inc. v. NLRB, 502 U.S. 527, 527 (1992) (holding that an employer did not commit an unfair labor practice by banning non-employee union organizers from the parking lot of its retail store). 67. Compare Four B Corp. v. NLRB, 163 F.3d 1177, 1183 (10th Cir. 1998) (applying an exception to the general rule; that an employer can not bar union solicitation on its property if the employer has permitted similar activities by other nonemployee groups) and Riesbeck Food Mkts., Inc., 315 N.L.R.B. 940, 942 (1994), enforcement denied, 91 F.3d 132 (4th Cir. 1996) (discussing the discriminatory nature of a company's policy and practice to distinguish among various solicitations of its customers), with Sandusky Mall Co. v. NLRB, 242 F.3d 682, 683 (6th Cir. 2001) (rejecting the Board's discrimination theory). 68. See Sears, Roebuck & Co. v. San Diego County Dist. Council of Carpenters, 436 U.S. 180, 180 (1978) (discussing the differences between issues presented to the state court versus the NLRB). 2002] EMPLOYEE ORGANIZATION IN SILICON VALLEY

"property" under California law, the California court was not wrong to enjoin Hamidi from using it. States are largely free in the constitutional system to define "property," and California has in fact decided that owners of shopping centers and other property open to the world do not have a property right to keep out union organizers. 69 People who do not have a state property right to exclude trespassers do not have a federal labor law right to do so, either.70 However, Intel's internal e-mail system, unlike shopping centers, is not routinely open to people other than Intel employees. All of these problems could be avoided in the future if FACE messages were sent by Intel employees lawfully authorized to send messages over the network. Sending such messages would be presumptively protected by the National Labor Relations Act. Employees could not be disciplined for sending such messages, unless Intel could show some unusual disruption or harm, or that the system is not open to employee messages of any kind.71 An employer with an e-mail or bulletin board open to all employees could not have a rule forbidding union or organizational messages. E-mail systems and databases of addresses are held to be property in cases involving either "spain" (unsolicited bulk e-mail) or what is now called "hacking" (an unauthorized breach of system security). If an outsider accesses a data system and misappropriates information or causes harm, the owner may sue under the Computer Fraud and Abuse Act (hereinafter "CFAA).72 It has become common practice in such suits to include a separate claim under state trespass law.73 Hamidi probably did not violate the CFAA, since he did not damage Intel's data or acquire any for himself. However, his messages were a kind of unsolicited bulk e-mail or spam, and spam is so unpopular that courts are unlikely to be criticized when they hold that owners of networks are empowered to ban it. It has been argued that courts should allow people like Hamidi to send messages

69. See PruneYard Shopping Ctr. v. Robbins, 447 U.S. 74, 88 (1980) (holding that speech and petitioning in privately-owned shopping centers are protected). 70. See Bristol Farms, Inc., 311 N.L.R.B. 437,437 (1993) (allowing union supporters to picket peacefully in front of a store). 71. See, e.g., Wash. Adventist Hosp., 291 N.L.R.B. 95, 95 (1988) (ruling on a hospital communication system that was not open to any employee messages). 72. 18 U.S.C §§ 1030 (1999) (prohibiting fraud in connection with computers). 73. See, e.g., Register.Corn v. Verio, Inc., 2000 U.S. Dist. LEXIS 18846, at *2 (S.D.N.Y. Dec. 8, 2000) (demonstrating that plaintiffs who sue for hacking under the CFAA of 1986 also sue for trespass); Am. Online Inc. v. Nat'l Health Care Discount, Inc., 121 F. Supp. 2d 1255, 1277 (N.D. Iowa 2000) (reaffirming that plaintiffs simultaneously sue under the CFAA and make claims for trespass of chattels); Ebay, Inc. v. Bidder's Edge, Inc., 100 F. Supp. 2d 1058, 1069 (N.D. Cal. 2000) (offering another example of a plaintiff suing under CFAA and also for trespass). 510 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3 to employees via e-mail.74 It would be a surprise if such access was attained by former employees or union organizers. Recall that employees with access to internal networks may send such messages, and are protected by federal law if they do. The most interesting legal issues raised by network groups and other "employee voice institutions" may arise in the long term. It is conventional to assert that labor law recognizes some harm to the public from unrestrained cartel effects and, on many issues, takes the form that it does in order to trim some of those cartel effects. This might be true of limitations on union secondary boycotts, application of antitrust laws to certain labor activities, and obligations that labor unions observe internal democracy and offer fair representation to individuals. If one really were faced with employee action that involved strictly "voice" effects without risk of cartel, one might expect greater scope for its action.

III. ETHNIC ORGANIZATIONS AND NETWORKS

The popular picture of the isolated Silicon Valley engineer, while not totally inaccurate, turns out to be somewhat ethnocentric. 5 Silicon Valley's large immigrant population makes extensive use, not only of informal ethnic ties, but also of formal, ethnic-based network organizations.76 About thirty percent of the Valley's high-tech workforce was foreign- born in 1990, about one-third of all scientists and engineers. Of these, almost two-thirds were Asian, the majority Chinese or Indian. 78 By contrast, half of this proportion of foreign-born scientists and engineers works in Massachusetts and Texas.7 9 Chinese and Indian engineers run one-quarter of the region's technology businesses started since 1980, in companies that collectively account for more than $12.5 billion in annual sales and 46,290 jobs.80 AnnaLee Saxenian demonstrates the importance of formal

74. See Martin H. Malin & Henry H. Perritt, Jr., The National Labor Relations Act in Cyberspace: Union Organizing in Electronic Workplaces, 49 U. KAN. L. REv. 1, 38-40 (2000) (asserting that it has been argued that people have a right to send unsolicited e- mails). See also The Long Arm of Cyber-Reach, 112 HARv. L. REV. 1610, 1622-24 (1999) (comparing Hamidi's e-mails to protected, in person, distribution of information). 75. See, e.g., BELLAH ET AL., supra note 2. 76. Here I use "network" in the older sociological sense, and not to imply use of computer-mediated communication. 77. See ANNALEE SAxENIAN, SILICON VALLEY'S NEW IMMIGRANT ENTREPRENEURS 11- 26 (1999). 78. Id. 79. Id. 80. Id. 2002] EMPLOYEE ORGANIZATION IN SILICON VALLEY organization in this economic success story. 81 She lists thirteen formal professional associations of Chinese or Indian engineers that foster networking and support along fairly narrow ethnic lines. 2 For example, there are distinct associations of engineers who originate from Taiwan who speak Mandarin at meetings, from Hong Kong who speak Cantonese, from mainland China, and Chinese engineers who prefer to speak English. 3 Another association was named The Indus Entrepreneur to include Pakistanis, Bangladeshi, and Nepalese; however, its members are almost all Indian. 4 These associations put professionals in touch with role models and sources of venture capital, serve as important sources of information about market or job opportunities, and offer formal and informal lessons in the basic facts of entrepreneurship, management, and English communication. 85 If unions may seem a little too old-fashioned to fit with high technology culture, what about ethnic-based network groups? They link the Valley community with the immigrant cultures of the last century. They also link the Valley's richest and poorest immigrants. The Mexican janitor and gardener are likely to be part of a network of relatives and friends, often from the same city in Mexico, who share living and economic arrangements and provide information on job opportunities.86 These ethnic networks were central in the successful organization of their union. 7 The Mexican networks, however, are rarely formal organizations with names and meetings, and they are not linked electronically. The only employee organization that can truly be said to be employed both by high- end and low-end contingent workers in the Valley is the ethnic-based network. For Chinese and Indian engineers, the chief economic aspect of the ethnic network is a path toward entrepreneurship. Entrepreneurship opportunities are also provided by the Mexican networks, though these networks also assist in union organizing." The engineer meets role models and sources of venture capital. The janitor advances by becoming a contractor himself or herself, or uses the network for an informal sales

81. Seeid. at 27-51. 82. Id. 83. Id. 84. Id. 85. Id. 86. See Zlolniski, supra note 9, at 2313 (affirming the existence of Mexican network groups). 87. See Saldafia, supra note 9 (indicating that network groups indeed affect union formation). 88. See id (discussing how both networks also perform some aspects of the labor market intermediaries that provide information matching employees to opportunities). 512 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3 business, or even an informal dentistry practice.89 The entrepreneurship path is crucial to understanding the groups as employee organizations and also to understanding patterns of immigration. It also increases the role of business ownership in a high-velocity labor market, in which employers do not offer lifetime careers and individual advancement normally requires exit from the firm and entry into either new employment, self-employment, or business ownership. Employees in such a high-velocity labor market can outperform employees in a traditional internal job tournament. 9° The promise of advancement outside the firm can be a powerful motivator of performance. For this to be true, the promise must be bound-as it is in the Valley-in a new implicit contract. The employee will leave his or her current firm with skill and experience that he or she lacked coming in and will be able to exploit that knowledge in the next workplace, or at the employee's own business. Based on this understanding of how the promise of entrepreneurship shapes employee organizations, this article suggests that other new employee organizations will take advantage of this promise. 91 In contrast to their entrepreneurship functions, ethnic-based networks of employees, while transcending particular employers, have not yet taken on the role of advancing employee interests as employees. One group that advances employee interests in the workplace is the Immigrants Support Network, which lobbies and organizes for changes in the treatment of immigrants on temporary visas. This article will address this group separately. Although dominated by Indian members, the group aspires to speak for immigrants more generally, so this article will treat them as a "new organization" rather than a "formal ethnic-based organization." Ethnic-based networks do not appear to have provided the basis for such informal protests as the Apple profit-sharing or IBM pension protests discussed earlier in this article. They do not go to bat for individual employees with grievances against particular employers or labor contractors. Though this must happen on an informal basis, no specific example was found. Are ethnic-based employee networks simply creatures of the recent immigrant experience that may fade away as immigrant groups become more acculturated? Or are they prototypes for new forms of employee organizations that can be used by other groups of employees-not merely new immigrant groups, but other groups of employees? Will such networks take on any additional functions of new labor market intermediaries: information, job matching, or possibly benefits? Will they

89. Zlolniski, supra note 9, at ch.2. 90. See HYDE, supra note *. 91. See generally HYDE, supra note * (discussing the role of entrepreneurship in civil rights or antidiscrimination strategies for high-velocity labor markets). 2002] EMPLOYEE ORGANIZATION IN SILICON VALLEY remain best suited to advancing job exit through entrepreneurship? Will they ever take on representational functions for employees? While it is too soon to answer any of these questions definitively, tentative answers may be gleaned from examining two other kinds of employee groups: formal ethnic caucuses organized by, or with the cooperation of, large firms, and women's networks.

IV. FORMAL ETHNIC CAUCUSES FOR LARGE EMPLOYERS

While no Silicon Valley employer has a union of technical personnel, several encourage formal caucuses. For example, Apple Computer has a formal Multicultural Alliance including Asian, African-American, Hispanic, Lesbian and Gay, Professional Women, and Vietnamese caucuses. These caucuses are defined almost exclusively as vehicles for the advancement of individual group members within firm hierarchies. They play little or no role in linking women, Latino, Asian or African- American employees across firm boundaries. Their inability to perform the latter function impedes their utility in the emerging high-velocity labor market. The basic model of an employee identity caucus is familiar.92 "The issue that African American network groups deal with is, broadly speaking, the acceptance, comfort, and career achievement of Black employees. 93 The same might be said of other network groups. Their main activities include mutual support, networking, exchange of information, particularly regarding managerial vacancies, and advice. Within a sample of Fortune and Service 500 companies responding to a survey, twenty-nine percent had network groups-typically African-American, women's, or both. Groups are often initiated by managers who are not part of the group, and who seek assistance in recruiting, or assistance for some other purpose, such as reviewing an affirmative action plan or planning Black History month. The group then stays together to pursue its own objectives. Groups tend to be dominated by professional or managerial employees even when nominally open to others. 4 They almost never make formal demands or requests of management. When managers express concerns

92. See generally RAYMOND A. FRIEDMAN & DONNA CARTER, AFRICAN AMERICAN NETWORK GROUPS: THEIR IMPACT AND EFFECTIVENESS (1993) (relying on phone interviews, surveys, and field visits to analyze the impact of African American Network Groups and make recommendations on the capabilities of such groups and how they may be formed); Raymond A. Friedman, Defining the Scope and Logic of Minority and Female Network Groups: Can Separation Enhance Integration? 14 RES. PERSONNEL & HuM. RESOURCES MGMT. 307, 308-12 (1996) (addressing the benefits of employee network groups devoted to furthering the interests of women and minorities in management positions). 93. FRIEDMAN & CARTER, supra note 92, at 3. 94. Id. at 16. 514 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3 about the formation of such groups, they fear that the groups will be confrontational and union-like. Groups are sensitive to this fear and avoid confrontation or demands. This is certainly true of the Silicon Valley groups. The sole categorical exception found through database news searches and personal discussions was gay and lesbian caucuses, which typically request that corporate benefits programs be opened to nonmarried domestic partners. When this request is granted, such groups generally make no further demands and revert to a more typical social-and-support group.95 Sometimes African-American or women's caucuses, particularly when asked to review an affirmative action plan, may suggest numerical goals. They may also request changes in formal job posting, evaluation, mentoring programs, sensitivity training and cultural awareness programs. Still, the human resources managers surveyed by Friedman and Carter ranked "influence policies" the second-lowest among potential functions actually accomplished by network groups. Thus, while almost no problems or downside can be identified in the existence of identity caucuses, their positive role seems largely limited to career and psychic support for their particular members. Silicon Valley firms often have active caucus systems that largely replicate this pattern of corporate loyalty and devotion to career advancement. For example, in March 1996, Apple Computer had a formal Multicultural Alliance consisting of six employee groups (Asian, Black, Hispanic, Lesbian and Gay, Professional Women, and Vietnamese).96 A "senior specialist, multicultural program" publicized their efforts and acted as liaison with management. Christian fellowship and Jewish cultural groups were, at that time, functioning outside the Alliance. All of these groups were "identity caucuses." The human resources specialist in 1996, Toni Tomacci, told me laughingly: "There are no Employees for a Democratic Apple, or for Higher Bonuses." Groups, funded mostly by individual contribution, got e-mail boxes, meeting places, and permission to plan public events. All of their activities consisted either of "social, networking" functions or of helping the company use diversity for business success, such as tapping Asian or Latin American markets. Groups have persuaded Apple to fund conferences of professional women or mentoring programs at historically black colleges. Particularly when planning conferences, groups may link up informally with counterpart caucuses at Hewlett-Packard or Xerox, but nobody knows of any links with larger community or political organizations. At least one group formally links identity caucuses, though it still

95. Maureen Scully, Remarks at the Spring Meeting of the Industrial Relations Research Association (1997). 96. Interview with Toni Tomacci, Senior Specialist, Human Resources, Apple Computer, (March 1996). 2002] EMPLOYEE ORGANIZATION IN SILICON VALLEY

excludes larger community or political organizations. A National Hispanic Employee Association (NHEA) is active in the Bay Area and elsewhere. The NHEA held Second Annual Breaking Barriers Awards Conference in San Jose, California, on March 15, 1996. However, the group is not even the beginning of an "associational union".97 Rather, it describes its "mission and vision.., to promote the upward mobility of Hispanics in the corporate and public sectors." 98 Indeed, the conference, sponsored by Silicon Graphics, Sun Microsystems, Apple Computers, AT&T, and similar employers, was largely aimed at high school and college students and advised them regarding self-presentation and career planning. The mission was clearly to facilitate the inclusion of Hispanics in existing corporate culture, and transformation of that culture only to the extent of including more Hispanics. This discussion is not intended to denigrate the importance of these goals or the efficacy of the NHEA's efforts. Rather, it is intended to question the disjunction between the corporate world uniformly invoked in the literature and the communications of employee network groups like NHEA or the Apple caucuses, and Silicon Valley's high-velocity labor market. The literature and speeches at the conference invoked images of a world of loyalty to the corporation, advancement through internal promotion ladders ("upward mobility"), and the concept of fitting into an existing corporate culture. There was no reference to a high-velocity labor market involving frequent job changes, startups, informal exchange of information and know-how sharing among firms. The intense devotion of identity caucuses and their spokespersons to the world of corporate loyalty, careers, and promotions, may show that world to be in better shape than this article too quickly assumes. On the other hand, it may be that there is a kind of deeper affinity between network groups and the hierarchical corporation, with internal labor markets, that may impede the network groups' recognition of newer and more vital ways of operating labor markets. Employee network groups, if they link with groups at other firms and community groups, could play a more important role in future labor markets than they currently play. They could take on some of the intermediary functions previously discussed, such as job matching and information, benefits delivery, and the additional functions of political and community advocacy. Such linked organizations could play a larger role in remedying problems of discrimination.99 It will be necessary, however, for

97. See generally HECKSCHER, supra note 23 (writing on the history of formal labor unions in the United States and the current and future role of associational unions in shaping the American workplace). 98. Id. 99. See generally HYDE, supra note * (exploring economic and legal issues inherent in 516 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3 these organizations to transcend a model of loyalty to hierarchical corporations with career ladders, and imagine functioning in high-velocity labor markets. Unless this occurs, it is hard to imagine any interesting consequences of corporate-oriented ethnic caucuses.

V. WOMEN'S NETWORKS

Women's networks will provide an interesting test over the next few years of the potential for identity-based groups to cross the boundaries of the firm and perform informational or other economic functions in a high- velocity labor market. The record on the role of women's networks to date is rather sparse. Silicon Valley's technology firms simultaneously offer unusual opportunities and obstacles to the careers of women as a group. Women are commonly chief financial officers or chief counsels, as a result of the formation of new companies, recent hiring, and a meritocratic culture. Obstacles revealed by surveys include balancing punishing work schedules with family obligations, and the pressure to conform to a masculine atmosphere. 100 A few formal network organizations of women transcend firm boundaries. Like the Chinese and Indian organizations studied by Saxenian, these networks are particularly oriented toward facilitating entrepreneurship through access to capital and other technical support.'01 The best-known is the Forum for Women Entrepreneurs, which claims through programs and networks to have assisted women in raising over $600 million in starting businesses in 2000.102 They run formal programs in which women entrepreneurs teach small groups of potential entrepreneurs.0 3 Other formal groups include Silicon Valley Webgrrls, San Francisco Women on the Web, and Women in Multimedia.' 4 These groups are known entirely by self-description and journalism. There are no ethnographies of these formal or informal organizations, and they must involve very few women. They never came up in interviews with women in the Valley, despite my questions on this topic. For women who plan to remain employees, or to be self-employed but not an entrepreneur, and who seek support in career planning or advocacy, a high-velocity labor market). 100. WOMEN OF SILICON VALLEY, UNFINISHED BUSINESS: WOMEN IN THE SILICON VALLEY ECONOMY 15-22 (2001), available at http://www.womenofsv.org. 101. See SAXENIAN, supra note 77. 102. WOMEN OF SILICON VALLEY, supra note 100, at 14. 103. Id. 104. See Karen Solomon, Girls' Night Out: Women Executives Network on Their Own Terms, INDUSTRY STANDARD, Aug. 7, 2000, available at http:l/www.thestandard.comlarticle/0,1902,17325,00.html. 20021 EMPLOYEE ORGANIZATION IN SILICON VALLEY or action for better child care, there seem to be few formal organizations. Women's career planning, like men's, often involves informal networks. A typical informal network consists of people one knew from former jobs, neighbors, and other social acquaintances. Naturally, people hear about jobs, new products and so on through whatever informal networks they employ, but no evidence was found to suggest that informal networks of women play any special role. It is too soon to conclude that women's networks will function similarly to any other kind of network, such as Chinese engineers, or website designers who once worked together. There are some intriguing hints in the literature that they may play unique roles. For example, the Stanford Project on Emerging Companies (hereinafter "SPEC") has reported the surprising fact that start-ups with high initial representation of women in core technical and scientific roles will have significantly fewer administrators in subsequent years. In the typical firm, about twenty-five percent of these roles were filled by females in its founding year. Such a firm would have only eighty percent as many full-time administrators in subsequent years, as compared to an all-male firm.'05 The numbers were large, and held constant across different variables. Why should this be? SPEC suggests a number of possible explanations. For instance, one such justification is that men like having titles."0 6 But in more recent work, SPEC explores the role of network ties."'7 Women are common in start-ups typically when they enter through network ties to the founders, and a firm with such network ties has richer informal control and needs fewer formal administrators 0 s Surveys reveal that women in high-tech fields attribute unusual importance to the role of networks and mentors in advancing their careers.'09 In short, there is much to learn about informal networks in high- velocity labor markets and any gender differences in their use.

105. See James N. Baron et al., Building the Iron Cage: Determinants of Managerial Intensity in the Early Years of Organizations, 64 AM. SOCIOLOGICAL REV. 527, 527-47 (1999) (examining how founding conditions shape the growth of management and administration in a sample of young technology companies in California's Silicon Valley). 106. Id. at 533. 107. See James N. Baron et al., Gender and the Organization-BuildingProcess in Young, High-Tech Finns, in THE NEW ECONOMIC SOCIOLOGY: DEVELOPMENTS IN AN EMERGING FIELD (forthcoming 2003). 108. Id. 109. WOMEN OF SILICON VALLEY, supra note 100, at 22 (reporting that forty-eight percent of women employed in the technology sector agree strongly "that developing networks and meeting the right people" are important to advancing in their career). 518 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3

VI. NEW ORGANIZATIONS FOR THE MOBILE WORKFORCE: WORKING PARTNERSHIPS, WASHTECH, WORKING TODAY, IMMIGRANTS SUPPORT NETWORK

New organizations have been created for mobile, high-tech workers, partly for those workers' sake, and partly because existing unions and some foundations regard them as potential prototypes for groups serving other contingent workers. These new organizations have attracted some of the most thoughtful, imaginative, and energetic people interested in labor issues, though their success has been limited to date. Particularly active in Silicon Valley are Working Partnerships, which is affiliated with the South Bay Labor Council AFL-CIO, and the Immigrants Support Network. Two other particularly interesting organizations are not yet active in the Valley. They are included here because the Valley would be an obvious next location for them if they achieve success. These are the specialized projects of the Communication Workers, which is aimed at employees of Microsoft in Seattle and IBM in upstate New York, and Working Today of New York City.

A. Working Partnerships/SouthBay Labor CouncilAFL-CIO

The union movement in the Valley has chosen some unusual forms through which to organize low-wage workers, particularly temporary and contingent workers. Working Partnerships USA (hereinafter "Working Partnerships"), founded in 1995, is affiliated directly with the Council, rather than any particular affiliated union."0 The group's title is intended to convey a desire to build bridges between organized labor and nonunion workers."' Working Partnerships has issued a series of research reports on the Silicon Valley economy, work organization, and housing 1 2 It has organized political campaigns around "living wage" ordinances enacted to raise labor standards in the contracts of local governments, and on behalf of

110. Telephone interview with Amy B. Dean, President, Working Partnerships USA, Oct. 10, 2001. 111. Id. 112. See Chris Benner & Amy Dean, Labor in the New Economy: Lessons from Labor Organizing in Silicon Valley, in NONSTANDARD WORK: THE NATURE AND CHALLENGES OF CHANGING EMPLOYMENT ARRANGEMENTS 361 (Franqoise Carr6 et al. eds., 2000) (discussing the significant components of a new model of employee representation conceived at the Temporary Worker Employment Project); Aaron Bernstein, Down and Out in Silicon Valley, Bus. WK., Mar. 27, 2000, at 76 (reporting on the plight of low-wage Silicon Valley workers left behind during the economic boom); see also Silicon Valley@Work, at http://www.atwork.org (last visited Jan. 24, 2002) (using a website forum to organize for justice in Silicon Valley, especially with regard to low wage workers and affordable housing). 20021 EMPLOYEE ORGANIZATION IN SILICON VALLEY favored candidates 13 It also runs leadership training courses for members of neighborhood associations in San Jose and other community leaders.114 Most intriguingly, Working Partnerships opened its own temporary help agency, hoping to compete with existing temp agencies while maintaining high labor standards 1 5 In October 2001, it was placing forty or fifty individuals a week for 1,200 hours on average.116 Its members work both in entry-level positions such as office temps, receptionists, and administrators, as well as systems administrators,1 7 network administrators, and cablers outsourced to Pacific Gas & Electric. The staffing agency incorporates training and upward career mobility." s It offers training in software skills and programming languages like Java and Juniper; A+, a standard training in hardware literacy; and Cisco and other certifications for network administrators and support staff."9 The unique supposed advantage of a union-affiliated staffing agency is this combination of job placement, training, and employee mobility. In contrast, junior colleges offer some training, but not linked to specific jobs. 120 Employers also offer some training, but rarely for mobile employees.121 Most staffing agencies offer little or no training.122 The theoretical and practical advantage of an employee-run organization is the combination of all these benefits in one place. Working Partnerships' employees currently are enrolled in a Kaiser health plan for mobile employees during the time that they are working.'23 The next project will be the creation of a trust fund, based on the model used by construction unions, into which employers will pay pennies per hour worked.124 The objective will be to permit some coverage of employees during the periods when they are between placements.125 All of these initiatives reflect a strategy of positioning worker organizations in larger political and geographic communities, not merely inside traditional firms, reflecting the decline in internal labor markets and one-firm careers. Thus, the political and community activity is not merely a device to organize a traditional union. It is a new model of union organizing that, if successful, will result in a very different type of union.

113. See sources cited supra note 112. 114. Id. 115. Id. 116. Telephone interview with Amy Dean, supranote 110. 117. Id. 118. Id. 119. Id. 120. Id. 121. Id. 122. Id. 123. Id. 124. Id. 125. Id. 520 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3

For example, its training courses and self-directed tutorials are administered jointly with the Black Chamber of Commerce. 126 It has warm, though less formal, ties with the Hispanic and Vietnamese Chambers of Commerce.' 27 Amy Dean pointed out how unusual these business ties are for a union. Every Labor Day, Working Partnerships, like other unions, sends representatives to speak at over eighty churches and synagogues, where congregants are asked to sign cards supporting its projects.1lz While this publicity technique is not unique to Working Partnerships, it takes on a different meaning in an organization devoted to community advocacy as much as worker representation. American unions have been organized around specific industries, or specific crafts, but there is little precedent for unions organized around regions and communities. Nor can one point to successful unions organized around a way of working, such as temporary or contingent, as opposed to a craft or job description, such as carpentry or waitressing. Working Partnerships draws on imaginative recent thinking on new ways to organize unions.' It combines elements of an associational union, 3 ° a regional craft union,"' and a union of low-income service workers, organized around their work organization as opposed to craft or industry. 3 2 Other scholarship supports Working Partnerships' union-run job referral system133 and training for workers in dead-end jobs to create career paths that lead out of those jobs. 34 A project this ambitious cannot be expected to achieve instant success. It will take time to sort out

126. Id. 127. Id. 128. Amy B. Dean, Remarks to the National Press Club, Presentation to National Conference on the Next Agenda, organized by the Campaign for America's Future (Feb. 28, 2001), available at http://www.atvorkorg, at 10. 129. Much of this scholarship is reviewed in Katherine V.W. Stone, The New Psychological Contract: Implications of the Changing Workplace for Labor and Employment Law, 48 UCLA L. REV. 519, 631-51 (2001) (suggesting methods to address problems of insecurity, unfairness, and injustice arising from the temporary nature of modem employer-employee relations). 130. See generally HECKSCHER, supra note 23 (discussing the history of formal labor unions in the United States and the current and future role of associational unions in shaping the American workplace). 131. See generally DOROTHY SUE COBBLE, DISHING IT OUT: WAITRESSES AND THEIR UNIONS IN THE TWENTIETH CENTURY (1991) (discussing the history and controversies surrounding the waitress as a craft unionist). 132. See Howard Wial, The Emerging Organizational Structure of Unionism in Low- Wage Services, 45 RUTGERS L. REv. 671, 692-93 (1993) (describing a model whereby low- wage workers are organized geographically along loose occupational lines). 133. See Eileen Silverstein & Peter Goselin, Intentionally Impermanent Employment and the Paradoxof Productivity, 26 STETSON L. REv. 1, 36-51 (1996) (proposing that unions act as employment agents for contingent workers). 134. See STEPHEN A. HERZENBERG ET AL., NEW RULES FOR A NEW ECONOMY: EMPLOYMENT AND OPPORTUNITY IN POSTINDUSTRIAL AMERICA 123-148 (1998) (exploring creation of career paths in the absence of tradional, hierarchical firms). 2002] EMPLOYEE ORGANIZATION IN SILICON VALLEY what works.

B. ImmigrantsSupport Network

More successful numerically is a group focused entirely on political representation, the Immigrants Support Network, representing mainly individuals working under "H-lB visas," or temporary visas issued to skilled individuals to fill supposed labor shortages. The entire program is controversial. 135 High-technology businesses around the country are often heavy users of such H-lB workers, nowhere more so than in Silicon Valley. The prototypical H-lB worker is a computer programmer from India (which provides almost half of all H-lBs), twenty-eight years old, 136 and earns around $45,000 a year. The visa program requires repeated Congressional reauthorization, a political process in which employers are dominant. In 2000, the year of the most recent reenactment, this requirement authorized 195,000 H-lB visas annually, each effective for six years. Visa holders must be sponsored by their United States employer, may not be self-employed, and may have difficulties changing jobs if their first job does not work out. Critics of the program charge that it depresses salaries for programmers.13 7 They claim that if employers were unable to hire programmers on short visas, they would be able to find plenty of programmers, including older programmers 3 and graduates of less-prestigious U.S. institutions. ' Employers would have to pay these individuals more than $45,000 a year.139 However, these criticisms have not been influential in the legislative process. Their proponents are not well-organized and have not received effective 140 support from professional associations or organized labor. H-lB visa holders' most frequent requests are that they would like a smoother transition to permanent resident status and increased mobility among jobs that are not with their original visa sponsor. 141 The Immigrants

135. See generally HYDE, supra note * (exploring economic and legal issues inherent in a high-velocity labor market). 136. U.S. GENERAL ACCOUNTING OFFICE, H-lB FOREIGN WORKERS: BErrER CONTROLS NEEDED TO HELP EMPLOYERS AND PROTECT WORKERS 4 (Sept. 7, 2000), available at http://www.gao.gov/new.items/heOO157.pdf (last visited Dec. 5, 2001). 137. The most prominent academic critic of the program is Professor Norman Matloff of the Computer Science Department at the University of California at Davis. His constantly- updated web page is an invaluable source of information. Norman Matloff, Debunking the Myth of a DesperateSoftware Labor Shortage, at http://heather.cs.ucdavis.edu/itaa.real.html (last visited Dec. 5, 2001). 138. Id. 139. See generally HYDE, supra note * (indicating that the allegations that the visa programs depress programmers' salaries may have merit). 140. Matloff, supra note 137. 141. Id. 522 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3

Support Network (hereinafter "ISN") was formed after the 1998 H-lB process by visa holders who felt they required representation independent of employer groups for these concerns. 42 It appears to be funded largely143 by successful entrepreneurs of Indian origin, such as Kanwal Rekhi. These entrepreneurs have no direct economic interest in reforming the H- lB program to meet visa holders' complaints. Rather, they fund the organization out of pride in Indian contributions to high technology and desire to assist Indian nationals in immigration. Still, it seems only a matter of time before the interests of H-lB visa holders conflict with some interest of the businessmen who fund their organization. Norman Matloff claims that the ISN spent heavily to employ expensive lobbyists in the 2000 legislative process, despite receiving contributions from only about 700 of the 15,000 individuals it claims as "members."' 44 It achieved some success in that 2000 lobbying initiative, obtaining extensions of H-1B visas for those whose permanent resident applications were being processed, and permission for employers to expedite the H-lB process for a fee. The ISN's next goal is abolition of the country quota on permanent resident cards. 141 Does the ISN belong in an article on employee organization? It is limited to immigrants, receives funds from employers, and largely lobbies around particular political interests. Does that mean that it is not an employee organization? Or is this the kind of employee organization of the future, linking employees across employers, moving from workplace to political advocacy, drawing strength from identity ties? The future of the ISN will tell us a great deal about the future of employee organization. Will it link with other employee groups, such as the South Bay Labor Council? The AFL-CIO's position on immigration has recently changed sharply. It no longer opposes immigration as a threat to current U.S. jobholders. It now welcomes immigration and seeks to organize immigrants. Consistent with that position, it did not oppose extension of the H-1B program. However, neither the AFL-CIO, nor any of its affiliates, has made efforts to organize H-lB workers. Could the ISN and organized labor such as Working Partnerships

142. Id. 143. See Aseem Chhabra, Rekhi's Visa Stand Raises a Ruckus, at http://www.rediff.com/news/2001/may/l0usspec.htm (last visited Dec. 3, 2001) (reporting that Mr. Rekhi, with a reported net worth of $500 million, has recently joined the ISN board). 144. Posting of Norman Matloff, Department of Computer Science, University of California, Davis, [email protected], to Age Discrimination/H-lB Visa Mailing List (May 7, 2001) (copy on file with author). 145. Raj Jayadev, Look Who's Organizing - High-Level Tech Workers Form New Labor Group, at http:llwww.pacificnews.org/contentlpns/2001/aug/0815techlabor.html (last visited Dec. 5, 2001). 2002] EMPLOYEE ORGANIZATION IN SILICON VALLEY work for a common cause? Such a joint program might seek ease of immigration, high employment standards, and unrestricted mobility for all workers, including immigrants. 46 In such a hypothetical alliance, these labor organizations would be turning their back on older programmers allegedly displaced by immigrants. However, this concern might not be an obstacle to such an alliance, in part because there are questions about the size of the group of displaced older programmers. Organized labor groups might well conclude that its future lies with newer immigrants rather than with older IT professionals who never affiliated with labor organizations. Could the ISN, either by itself or with assistance from organized labor, become an advocate for immigrants in disputes with their employers or sponsors, in addition to becoming a lobbyist in Congress? Could the ethnic pride and energy of Silicon Valley's immigrant communities fuel successful organizing for programmers, as it has for janitors?

C. Communication Workers ProjectsAimed at IT Workers

The Communication Workers of America (hereinafter "CWA"), which represents many workers at older telephone companies, has set up two innovative projects to advance the interests of information professionals: the Washington Alliance of Technology Workers (hereinafter "WashTech") in Seattle, and the Alliance@IBM in Endicott, New York. These projects advance workers' interests without acting as their legal bargaining representative, and thus have been described as "virtual unions."' 47 WashTech is actually Local 37083 of the Communication Workers, 4 but does not yet function as a bargaining representative for its members.1 1 WashTech focuses on the workers at Microsoft who are nominal employees of temporary help agencies. In the mid-1990's, the Ninth Circuit found that some of these individuals were legally employees of Microsoft.' 49 Microsoft has changed its practices since then, and presumably most individuals whom it classifies presently as agency employees would also be found to be agency employees in law. WashTech provides lobbying, training, and informational services to these temporary

146. In other words, ending constraints on all workers' free mobility, such as covenants not to compete and some aspects of trade secrets law discussed in HYDE, supra note *,chs. 2 & 3, is consistent with abolishing those constraints limited to the H-lB visa program. 147. Wayne J. Diamond & Richard B. Freeman, Will Unionism Prosper in Cyber-Space? The Promise of the Internetfor Employee Organization,NBER WORKING PAPER 8483, at 15- 16 (October 2001), available at http://papersnber.org/papers/W84483. 148. Virginia L. DuRivage, CWA's Organizing Strategies: Transforming Contract Jobs into Union Jobs, in NONSTANDARD WORK,supra note 112, at 377-391. 149. Vizcaino v. Microsoft, 120 F.3d 1006, 1006 (9th Cir. 1997), cert. denied, 522 U.S. 1098 (1998). 524 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3 employees. While WashTech did not initiate the Vizcaino litigation, it kept its members informed about it. It has lobbied in the State of Washington for access to personnel records and for favorable employment standards treatment for agency employees, and has offered training classes in software programming and other skills. It has requested bargaining recognition, albeit unsuccessfully, from at least four temporary agencies. However, like Working Partnerships, WashTech aspires to anchor its organization in an industry and geographic area, not just in a collective bargaining relationship. In the words of Larry Cohen, executive vice president of the CWA, at WashTech "what sustains people is not collective bargaining, but creating a community of people. '150 Like the ISN or any web-based organization, it has many more individuals visiting its webpage and joining for specific actions than it has dues-paying members. As of June 2001, there were only 225 such members. 5 Alliance@IBM, another CWA affiliate, was formed in the Summer of 1999 in response to the electronic protests against IBM's proposed reductions in pensions, discussed earlier in this article.152 Its main constituency appears to be long-time IBM employees concerned about outsourcing and downsizing. 53 Alliance@IBM does not appear to have reached out to the temporary employees served by WashTech, or to immigrants on H-1B visas served by ISN. The interests of both of those groups may conflict with those of the career IBM employees. Perhaps at some point the CWA will have to confront this conflict, though at the moment it is free to organize opportunistically, which RedHerring magazine calls "the venture capital approach to unionizing."1 54 The CWA is not yet active in Silicon Valley, but this would be an obvious next arena if it feels that either WashTech or Alliance@IBM has been successful.

D. Working Today

Working Today was founded to represent mobile individuals, mostly self-employed, who lack ties to a single employer. So far, it has been funded more by foundations than by its members. Some of its founders, such as Columbia sociologist Herbert Gans, contemplated a group that would seek to convert such contingent work into more traditional, stable work, but the group eventually followed the path of other organizations

150. Steven Greenhouse, Five Questionsfor Larry Cohen: Renewing a Union in the New Economy, N.Y. TIMES, Dec. 24, 2000, at BU4. 151. See Jonathan Littman, Phone union attempts dial-up Internet connection, REDHERRING (June 19, 2001), at http://www.redherring.com/insider /2001/ 0619/ 420019642.html. 152. Kahn, supra note 36. 153. See Littman, supranote 151. 154. Id. 20021 EMPLOYEE ORGANIZATION IN SILICON VALLEY reviewed in this section. Working Today accepts the appeal of self- employment and a mobile workforce, and now hopes to learn what kinds of benefits appeal to such a group, how to market benefits to them, and what other kinds of economic and political representation they need. 55 Its first pilot project, launched in 2001, seeks to market health insurance to website designers in New York City. The group has designed what it believes will be an attractive package of portable benefits, even though they are still fairly expensive. It remains to be seen whether a participatory or democratic employee organization will have any advantage over the private sector in the marketing of benefits such as health insurance.

VII. SUMMING UP: THE FUTURE OF EMPLOYEE ORGANIZATION IN HIGH- TECHNOLOGY

The brutal truth is that Silicon Valley attained its present technological and economic preeminence without formal employee organization. It is difficult to see why this will change any time soon. The nascent organizations discussed in this article-unions, networks, caucuses, new organizations-may be interesting straws in the wind, and may present intriguing legal or economic issues. However, none of these organizations has achieved much importance. Put another way, employee organizations only come up in interviews when the interviewer specifically asks about them. Employee organizations will never be discussed when the conversation concerns where an employee worked, was an Human Resources director, or counselor. Many managers believe that employee organization is fundamentally incompatible with high-technology employment, in particular with the flexibility and energy demanded of and received from employees. When asked whether Cisco had formal women's or Black caucuses like Apple's, Michael LaBianca, the lawyer who handles employment matters, laughed and said, "People are too busy to think about anything else. ' 156 Managers would nearly all interpret a story like the Individuate story as one in which naive attempts at formal organization actually converted minor discontent into a serious problem. It is difficult to be optimistic about the new groups that hope to get employees into formal organizations by meeting employee desires for training (WashTech), benefits (Working Today), or job placement (Working Partnerships). Private organizations already exist to meet these needs, and there seems no practical or theoretical reason why an employee organization should be able to provide better courses, or insurance plans, than private companies.

155. Sara Horowitz, New Thinking on Worker Groups' Role in a Flexible Economy, in NONSTANDARD WORK, supra note 112, at 393-398. 156. Interview with Michael LaBianca, Cisco, in San Jose, Cal. (June 2001). 526 U. PA. JOURNAL OF LABOR AND EMPLOYMENT LAW [Vol. 4:3

It is easier to imagine a role for employee organization in the low- wage sector of the Silicon Valley labor market: the office temps, product testers, and perhaps low-level programmers. The model that would work best for them, however, may not be any of the new organizations designed for mobile workers. Rather, low-wage employees need a good, old- fashioned construction-type union. The employer on a unionized construction job pays contractual amounts, per hour worked, into trusts, which are jointly administered by the union and the employers, and pay health, vacation, and retirement benefits. This is the best institution ever designed for providing income stability, insurance, and retirement to what is really a very contingent work force. It is puzzling that so much energy has been put into designing new organizations for a mobile workforce, when United States law and practice already provide an excellent blueprint. Part of the problem with adapting the construction union model to the Silicon Valley market is the fact that construction work is a kind of contingent work, but one in which unions make earnings less contingent. Many construction workers work for many different contractors and on many jobs. Despite this uncertainty, construction workers are not usually included in discussions of "contingent" work. The difference is the union. Construction workers who are represented by a labor union typically have health insurance (87.1%), paid vacations, and a pension (67%).57 Construction employees who are not represented by a union normally have no health insurance (only 41.4% get it through work) or pension (only 22% have one).158 While the oldest construction unions involve highly-skilled employees, this structure has also been employed by laborers and others who have less formal training. Also, construction unions often operate hiring halls that provide labor to contractors on request. While this can facilitate administration of the benefits plans, it is not necessary for their success. Therefore, there is no obvious reason why office temps and other mobile, less-skilled employees could not be organized into unions that would administer benefits and might, or might not, take on training and job placement functions. Why do temps not have construction-type unions? We cannot be confident that temps would not join unions if they could. Of course, many people feel that it is generally true of all United States unions that the legal obstacles to their organization are so serious as to depress levels of organization below existing demand.5 9 Unions for mobile temps face

157. CENTER TO PROTECT WORKERS' RIGHTS, THE CONSTRUCTION CHART BOOK: THE U.S. CONSTRUCTION INDUSTRY AND ITS WORKERS, charts 3, 26, 27 (2d ed. 1998). 158. Id. 159. See generally PAUL C. WEILER, GOVERNING THE WORKPLACE: THE FUTURE OF LABOR AND EMPLOYMENT LAW (1990) (asserting that U.S. labor laws discourage union organization); FREEMAN & ROGERS, supra note 18. 20021 EMPLOYEE ORGANIZATION IN SILICON VALLEY problems in addition to these general impediments to unionization. The chief specific legal impediment to the organizations proposed for mobile employees is the lack of authority in the National Labor Relations Board to certify a bargaining unit with multiple employers. Such bargaining does take place, but only with the voluntary consent of the employer, who is free to withdraw from multi-employer bargaining at any time except when new contract negotiations are actually underway. The NLRB has retreated from some extensions of this doctrine that impeded union organization among temporary help workers. For years, the NLRB took the position that including temporary-help employees in a bargaining unit with regular employees was a kind of multi-employer bargaining that required the mutual consent of the temporary help agency and the client firm. The NLRB will now certify a unit of all employees working for one client-regular employees and those jointly employed by the temp agency can bargain with the client. The NLRB will also certify a unit of all the temporary employees working at various locations but referred by the same temp agency, in order to bargain with that temp agency. It is possible that these changes will spark new union organizing among temporary help employees. However, they represent the outer limits of the NLRB's power. A unit of all the temps who work for many different clients is considered a multi-employer unit and, consequently, bargaining is voluntary for each client.160 Nor does the NLRB have authority to force any client or temp agency to bargain jointly with the employees whom they jointly employ. Further, a congressional amendment to the National Labor Relations Act to facilitate multi- employer units is not imminent. However, multi-employer bargaining for temps can come about through the employer's voluntary actions, which is not totally far-fetched in Silicon Valley. After all, Valley employers preferred that their cleaning contractors bargain with the Service Employees, rather than face potential product boycotts. Therefore, while the most realistic assessment of the future is that rates of union organization in Silicon Valley will continue to be very low, if organization increases, it will not be the new-fangled caucuses, associations, and networks. Rather, the construction union form will turn out to be adaptable to serve other individuals who work for many different employers.

160. See In re M.B. Sturgis, Inc., 331 N.L.R.B. No. 173, 1 (2000) (reaffirming the employer consent requirement for the creation of true multi-employer units involving separate user employers).